or service under conditions or prooaoie, compared to zero, population growth. This measures the added burden that population growth adds to developmental costs of health services. Both methods are further complicated by considering two "probable" rates of population growth. One assumes conditions of no change in level or trend of natality from that of the past decade. The other assumes conditions of a decline in natality brought about primarily through organized natality control efforts by the country. Finally, the measure most meaningful to leaders of LDC's is the difference between the actual requirements for the two "probable" rates of population growth. Stated another way, this difference is the savings in investment resulting from effective natality control (Table 7). TABLE 7 Health Service Requirements under Differing Conditions of Population Growth and Development Change in Level of Health Service 20 Years from Now Population Growth in Next 20 Years Zero Without Natality Control With Natality Control Savings Zero Increased as planned A D B E C F B-C E-F Requirements A and D are theoretical, if no population growth occurred. Requirement A has for simplicity been taken as zero in this paper. Requirement D is simply the initial population multiplied by the increase in service: population ratio desired. All requirements are understated to the extent that they do not include replacements necessary for manpower who die or retire or migrate or, in some cases, for facilities and equipment that wear out. In view of the magnitude of requirements and of the difficulties of predicting future retirement and especially migration rates, it has seemed adequate here to assume that neither mortality nor retirement will be significant before 1985 because of the young age distribution of doctors and nurses, and that no real improvement in levels of manpower can occur without developing working conditions that will minimize a "brain drain." Need to replace losses from mortality, retirement, and migration can be great; in the case of Taiwan replacement needs were estimated for the 20 years 1963-1983 to be 76 percent of the total supply of doctors (28, p. 68) in 1963. Requirements B and E are based upon population projections which assume mortality and natality declining at approximately the 1955-1965 rate (which for most LDC's means no change in natality rate). Requirements Cange by individual country was less then 10 per 100,000 and in nineteen was less then 5. Most of the decreases were in newly independent African countries. For the same time period, the increase in ratio of physicians to population for the United States was 14.4 percent of the 1955 level or 18 doctors per 100,000 population.